How does IPFS challenge our existing network?

Data issues

Since 2005, humanity has collected about 13 billion bytes of data. If this information were to be displayed on paper, it would take all the trees in the world to turn this data into books. In 2015 , this number was 7900 gigabytes. By 2020, this number will reach 40900, and this exponential growth of information will be consumed by the rapidly growing number of Internet users.

In Indonesia alone, 105 million Internet users are expected to grow to 125 million in the next four years. In the next four years, 20 million people will be online. And this is just for one country.

To complicate matters further, the difference between annual price declines for storage (40%) and bandwidth (26%) is almost twice as large.

All this means is that more people will be transmitting more data on the same channels. This will create congestion that cannot be solved by simply adding more hardware.

Congestion is just one problem, the internet today is a centralized system and naturally has a single point of failure. This point of failure can be abused to make the web inaccessible to an entire country. This is exactly what happened in Egypt during the January Revolution of 2011. When the president didn’t want to be overthrown, he decided to take down the internet that was powering the protests. Ultimately it didn’t help as far as we know, but the effect was clear, four major ISP providers cut the wires and the country was cut off from the world. The single point of failure here was that all communications between protesters were going in and out of Egypt, not directly between people, and when that door was closed, nothing worked anymore. Clearly, there is a problem, now let’s look at how businesses can benefit from it by investing in technology that provides a solution. This is where IPFS and blockchain come in.

IPFS

IPFS Interplanetary File System is a new Internet protocol, originally designed by JuanBenet in 2014, whose goal is to store data permanently, eliminate duplication on the network, and obtain addresses for information stored on network computers. It is now an open source project. Unlike HTTP, it is content-addressed, not location-addressed. IPFS points to the resource itself, rather than communicating by pointing to a location. It then takes this resource from anyone who has the data or part of it, creating a decentralized network.

So why should you care if the internet protocol you use is location-based? To do this, we can compare the two approaches.

HTTP works by mapping resources using location-based IP addresses. It uses them to point to a computer in a specific location that has the resource the client needs. For example, if you want to stream a 4k video on Youtube, your browser needs to look for that video on a server far away in a Google data center and have that stream come all the way to you. Actually, there is something called a Content Delivery Network (CDN) that can "bring the server" closer to you, but it may still be hundreds of kilometers away and is of little help in developing countries.

Imagine you're at a lecture with 100 other people, and you all watch the same video. What happens is that the video has to be fetched from the nearest node at Google, streamed to each student's laptop and repeated 100 times. Instead of students having the same copy of the video, sharing it with each other, we have a large amount of data being transmitted multiple times, over long distances. It's inefficient, but that's how HTTP works, and it's causing massive congestion issues on the Internet backbone.

Myth

Data stored on IPFS is persistent

Permanence and persistence are not the same thing. Let's be clear here, IPFS does not guarantee persistence, nodes on the network decide if they think it's worth keeping something on IPFS. What you can do with IPFS, though, is that you can create content, add it to the network, and then disappear without keeping any infrastructure running. Using cryptography, IPFS will serve your content as long as the network deems it valuable. The link to your unique content will always remain the same, but it's up to the user whether the network stores it or not. IPFS uses cryptocurrency to incentivize data storage, more on that later.

IPFS is built on the blockchain

Although it uses architectural elements similar to MerkleTrees, IPFS is not built on the blockchain, but is designed to work with existing blockchain protocols. As mentioned above, each file uploaded to IPFS will get a permanent address. Unlike IPFS, blockchain is not suitable for storing large amounts of data. Putting IPFS and blockchain together, you can store a lot of information on IPFS and put immutable permanent IPFS addresses into blockchain transactions. IPFS will provide a publicly accessible database, while the blockchain makes it publicly verifiable.

IPFS and Decentralization and Distribution

IFPS is a distributed file system that enables fast performance and decentralized data archiving.

Distributed means that all data processing/archiving is not done in the same place. Those distributed services can still be under the control of a single entity.

Decentralization means that no single entity controls all transactions.

IPFS is then distributed as data is stored across the network. Decentralization or ownership depends on the data and applications that manage this data.

IPFS and blockchain applications

IPFS is used to store publicly accessible data, while blockchain is used to verify addresses. This means two things.